Cancer progression is driven by an evolutionary process of genetic mutation, phenotypic variation, and selection. The genetic background of cancer cells is thus continuously changing. Genetic background will dictate the nature of future mutations that can be tolerated and selected for. The effects of specific new mutations will depend on the genetic background in which they occur. Thus the temporal order of mutations is expected to influence cancer phenotype. How the timing of mutation affects cancer phenotypes remains a largely unanswered question, however. Answering this provocative question (PQB5) is significant because cancer prognosis and effective personalized cancer treatment will depend not only on the nature of mutations present within a particular cancer, but also on the order in which they are acquired. Genetically engineered mouse cancer models have been vital in elucidating the genetic etiology of cancer, but have rarely been used to address PQB5. This is due to limitations inherent in commonly used genetic engineering methodology; when multiple mutations are created, the order of mutations is either unknown, is not controllable, or is not restricted to incipient cancer cells. A current barrier to progress is the availability of mouse cancer models that allow precise control over the timing and order of multiple genetic mutations. We propose to overcome this barrier by developing a novel mouse cancer model based on a lox-neo/stop-lox FlpO-ERT2 transgene. This transgene encodes an FlpO recombinase, ERT2 estrogen receptor fusion protein whose activity is tamoxifen inducible. The transgene is combined with tissue specific Cre transgenes, floxed alleles of the cancer initiating mutation, and frted alleles of the secondary mutation. Cre expression creates the cancer initiating mutation and removes the neo/stop cassette from lox-neo/stop-lox FlpO-ERT2, restricting its expression to initiated cancer cells. FlpO-ERT2 activity can then be induced at experimentally controlled times by tamoxifen administration, creating a secondary mutation by deleting frted gene alleles. This mouse model can be used to determine how 2 mutations occurring in a defined and controllable temporal order affect cancer phenotypes in vivo. We will use this mouse model to test whether the timing of Pten and Rb1 mutation alters prostate cancer phenotype. PTEN and RB1 mutation are common in human prostate cancer, with PTEN loss occurring early and RB1 loss occurring late. The late loss of RB1 is puzzling as its early loss is known to initiate other human cancers. It is unknown whether PTEN and RB1 loss cooperate to drive prostate cancer progression or whether the temporal pattern of mutation influences prostate cancer phenotype. We postulate that the timing of Pten and Rb1 mutation will alter prostate cancer phenotypes in vivo because the effects of these mutations are dependent on genetic background. Two specific aims are proposed to create the lox-neo/stop-lox FlpO-ERT2 allele, to use it to alter the timing of Pten and Rb1 mutation in the mouse prostate cancer model, and to characterize the effects of these mutations on prostate cancer phenotypes in vivo.